Enterobacter sakazakii and Salmonella in powdered infant formula Meeting Report

Transcription

1 Enterobacter sakazakii and Salmonella in powdered infant formula Meeting Report FAO Headquarters Rome, Italy January 2006 Advance pre-publication copy 2 May 2006 Note: This report is being released as a pre-publication copy to facilitate the provision of timely scientific advice to Codex Alimentarius and FAO and WHO member countries. However, please note that this document will undergo final copy-editing and formatting before publication. Whilst there will be no substantive changes to the content of the final published document, the format will be modified to comply with that of the FAO/WHO Microbiological Risk Assessment Series and there will be minor editorial changes.

2

3 iii CONTENTS Acknowledgements...v Contributors... vii Foreword... ix Abbreviations... xi Executive summary... xii Recommendations...xiv 1. Introduction and background Background Objectives Scope Micro-organisms other than Enterobacter sakazakii and Salmonella Epidemiological and microbiological aspects of Enterobacter sakazakii Epidemiology and Public Health Outbreaks Update since Update on microbiological aspects Characterisation of the micro-organism Isolation and identification Persistence in a desiccated state Pathogenicity and virulence Dose-response Conclusions Quantitative risk assessment model for Enterobacter sakazakii in powdered infant formula Introduction Overview of the risk assessment model Hazard characterization Exposure assessment Risk characterization Review of the model Conclusions Evaluation of risk management interventions for Enterobacter sakazakii in powdered infant formula Impact of control measures in the manufacturing plant Current industry practices Processing environment Dry mix ingredients Conclusions Preparation, storage and handling practices Consumer (hospital and home) practices Selected scenarios Relative risk - Results of modelling Conclusions Microbiological criteria for powdered infant formula Two-Class Attribute Sampling Plans for Enterobacter sakazakii in PIF...43

4 iv Sampling plans for Enterobacteriaceae Consideration of the cumulative impact of sampling plans through changes in production management and lot rejection Illustrative example of the cumulative impact of sampling plans through changes in production management and lot rejection Conclusions Salmonella in powdered infant formula Epidemiology and public health Epidemiology of Salmonella infections Recent outbreaks of salmonellosis associated with powdered infant formula Dose-response Update on microbiological aspects Risk reduction strategies Current industry practices Consideration of current microbiological criteria Preparation and use of PIF by consumers Need for risk assessment? Conclusions Response to Codex and Recommendations Response to Codex Other key outcomes Recommendations References...69 Appendix A - List of background papers and presentations...73 Appendix B - Data received in response to the FAO/WHO call for data...74 Appendix C - Review of risk assessment model...76 Appendix D - Impact of rehydration at high temperatures on other characteristics of PIF...79 Appendix E List of 45 cases of invasive Enterobacter sakazakii infections analysed...82

5 v ACKNOWLEDGEMENTS The Food and Agriculture Organization of the United Nations and the World Health Organization would like to express their appreciation to all those who contributed to the preparation of this report through the provision of their time and expertise, data and other relevant information. In particular the work of Greg Paoli and Emma Hartnett in developing the risk assessment model is acknowledged. Appreciation is also extended to all those who responded to the call for data that was issued by FAO and WHO and brought to their attention information that was not readily available in the mainstream literature and official documentation. The preparatory work and expert meeting convened to prepare this report was coordinated by the Joint FAO/WHO Secretariat on Risk Assessment of Microbiological Hazards in Foods (JEMRA). This included Sarah Cahill and Maria de Lourdes Costarrica in FAO, and Peter Karim BenEmbarek and Jørgen Schlundt in WHO. The secretariat was supported by Kaye Wachsmuth who also assisted in the finalization of the report. Publication of the report was coordinated by Sarah Cahill. The work was supported and funded by the FAO Food Quality and Standards Service and the WHO Department of Food Safety, Zoonoses and Foodborne Disease.

6

7 vii CONTRIBUTORS (MEETING PARTICIPANTS) EXPERTS Anna Bowen, Centers for Disease Control and Prevention, United States of America Robert Buchanan, Food and Drug Administration, United States of America Jean-Louis Cordier, Nestlé, Switzerland Alejandro Cravioto, ICDDR, B Centre for Health and Population Research, Bangladesh Sue Dobson, Australian Food Safety Centre of Excellence, Australia Sri Estuningsih, Bogor Agricultural University, Indonesia Jeffrey Farber, Health Canada, Canada Aamir Fazil, Public Health Agency of Canada, Canada Stephen Forsythe, Nottingham Trent University, United Kingdom Emma Hartnett, Decisionalysis Risk Consultants Inc., Canada Lisa Lefferts, Consultant, Lisa Y. Lefferts Consulting, United States of America Greg Paoli, Decisionalysis Risk Consultants Inc., Canada Alan Reilly, Food Safety Authority of Ireland, Ireland Hajime Toyofuku, National Institute of Health Sciences, Japan Bruno Coignard, Département Maladies Infectieuses - Institut de Veille Sanitaire, France Marcel Zwietering, Wageningen University, The Netherlands RESOURCE PEOPLE Kaye Wachsmuth, Public Health Microbiology Consultant, United States of America Verna Carolissen-Mackay, Codex Secretariat, Italy Jeronimas Maskeliunas, Codex Secretariat, Italy DECLARATIONS OF INTEREST Five of the 16 experts who participated in this meeting declared an interest in the topics under consideration: Dr Anna Bowen: Fifty per cent of her salary is covered by a commercial entity as part of a Cooperative Research and Development Agreement. While the company manufactures products for infants, it does not manufacture any infant formula products. Dr Jean-Louis Cordier: He is an employee of a manufacturer of powdered infant formula products. Dr Susan Dobson: She acted as a reviewer of a risk assessment on Enterobacter sakazakii in powdered infant formula for a commercial entity in Australia.

8 viii Dr Stephen Forsythe: His research related to isolation media for Enterobacter sakazakii and the virulence of Enterobacteriaceae is funded by two commercial entities, one of which manufactures powdered infant formula. Dr Marcel Zwietering: His unit at the Wageningen University performs scientific research which is financially supported by a company producing powdered infant formula. Those of Dr s Cordier, Forsythe and Zwietering were considered to represent potential conflicts of interest. Therefore, these experts did not participate in the final adoption of the conclusions and recommendations of the meeting.

9 ix FOREWORD The Members of the Food and Agriculture Organization of the United Nations (FAO) and of the World Health Organization (WHO) have expressed concern regarding the level of safety of food both at national and international levels. Increasing foodborne disease incidence over the last decades seems, in many countries, to be related to an increase in disease caused by micro-organisms in food. This concern has been voiced in meetings of the Governing Bodies of both Organizations and in the Codex Alimentarius Commission. It is not easy to decide whether the suggested increase is real or an artefact of changes in other areas, such as improved disease surveillance or better detection methods for micro-organisms in patients and/or foods. However, the important issue is whether new tools or revised and improved actions can contribute to our ability to lower the disease burden and provide safer food. Fortunately new tools, which can facilitate actions, seem to be on their way. Over the past decade, Risk Analysis a process consisting of risk assessment, risk management and risk communication has emerged as a structured model for improving our food control systems with the objectives of producing safer food, reducing the numbers of foodborne illnesses and facilitating domestic and international trade in food. Furthermore, we are moving towards a more holistic approach to food safety, where the entire food chain needs to be considered in efforts to produce safer food. As with any model, tools are needed for the implementation of the risk analysis paradigm. Risk assessment is the science-based component of risk analysis. Science today provides us with in-depth information on life in the world we live in. It has allowed us to accumulate a wealth of knowledge on microscopic organisms, their growth, survival and death, even their genetic make-up. It has given us an understanding of food production, processing and preservation, and of the link between the microscopic and the macroscopic world and how we can benefit from as well as suffer from these micro-organisms. Risk assessment provides us with a framework for organizing these data and information and to better understand the interaction between micro-organisms, foods and human illness. It provides us with the ability to estimate the risk to human health from specific micro-organisms in foods and gives us a tool with which we can compare and evaluate different scenarios, as well as identify the types of data necessary for estimating and optimizing mitigating interventions. Microbiological risk assessment can be considered as a tool that can be used in the management of the risks posed by foodborne pathogens, including the elaboration of standards for food in international trade. However, undertaking a microbiological risk assessment (MRA), particularly quantitative MRA, is recognized as a resource-intensive task requiring a multidisciplinary approach. Yet foodborne illness is among the most widespread public health problems, creating social and economic burdens as well as human suffering, making it a concern that all countries need to address. As risk assessment can also be used to justify the introduction of more stringent standards for imported foods, a knowledge of MRA is important for trade purposes, and there is a need to provide countries with the tools for understanding and, if possible, undertaking MRA. This need, combined with that of the Codex Alimentarius for risk-based scientific advice, led FAO and WHO to undertake a programme of activities on MRA at the international level. The Food Quality and Standards Service, FAO, and the Department of Food Safety, Zoonoses and Foodborne Diseases, WHO, are the lead units responsible for this initiative. The two groups have worked together to develop the area of MRA at the international level for application at both the national and international levels. This work has been greatly facilitated by the contribution of people from around the world with expertise in microbiology, mathematical modelling, epidemiology and food technology to name but a few. This Microbiological Risk Assessment series provides a range of data and information to those who need to understand or undertake MRA. It comprises risk assessments of particular pathogen-commodity combinations, interpretative summaries of the risk assessments,

10 x guidelines for undertaking and using risk assessment, and reports addressing other pertinent aspects of MRA. We hope that this series will provide a greater insight into MRA, how it is undertaken and how it can be used. We strongly believe that this is an area that should be developed in the international sphere, and have already from the present work clear indications that an international approach and early agreement in this area will strengthen the future potential for use of this tool in all parts of the world, as well as in international standard setting. We would welcome comments and feedback on any of the documents within this series so that we can endeavour to provide Member countries, Codex Alimentarius and other users of this material with the information they need to use risk-based tools, with the ultimate objective of ensuring that safe food is available for all consumers. Ezzeddine Boutrif Food Quality and Standards Service FAO Jørgen Schlundt Department of Food Safety, Zoonoses and Foodborne Disease, WHO

11 xi ABBREVIATIONS ATCC BAM BGA BPW cal CCFH CCP CDC CDLT cfu cm DFI DW EC EE ESIA ESPM ESSM EU FAO FDA g GHP GMP HACCP IDF ISO IU JEMRA kg l mg ml mlst MPN NGOs NICU o C PCR PIF TS TSA USDA VRBGA VRBLA WHA WHO XLD g American Type Culture Collection Bacteriological Analytical Manual (USFDA) Brilliant green agar Buffered peptone water Calories Codex Committee on Food Hygiene Critical control point Centers for Disease Control and Prevention (USA) Cytolethal distending toxin Colony forming unit centimetre(s) Druggan-Forsythe-Iverson (medium) Deionised water European Communities Enterobacteriaceae enrichment (broth) Enterobacter sakazakii isolation agar Enterobacter sakazakii chromogenic plating medium Enterobacter sakazakii screening medium European Union Food and Agriculture Organization of the United Nations Food and Drug Administration (USA) Gram(s) Good hygiene practices Good manufacturing practice Hazard analysis and critical control point system International Dairy Federation International Standards Organization International units Joint FAO/WHO Expert Meetings on Microbiological Risk Assessment Kilogram(s) Litre Milligram Millilitre(s) Modified lauryl sulfate tryptose (broth) Most probable number Non-governmental organizations Neonatal intensive care unit Degrees Celsius Polymerase chain reaction Powdered infant formula Technical Standard Tryptic soy agar United States Department of Agriculture Violet red bile glucose agar Violet red bile lactose agar World Health Assembly World Health Organization Xylose-lysine-deoxycholate agar Microgram

12 xii EXECUTIVE SUMMARY The 37 th Session of the Codex Committee on Food Hygiene (2005) requested the Food and Agriculture Organization of the United Nations (FAO) and the World Health Organization (WHO) to extend the scientific advice provided by the expert meeting on Enterobacter sakazakii and other micro-organisms in powdered infant formula held in Geneva in 2004 (FAO/WHO, 2004). Accordingly, a technical meeting was convened on E. sakazakii and Salmonella in powdered infant formula (FAO, Rome, January 2006) to consider any new scientific data and to evaluate and apply a quantitative risk assessment model for E. sakazakii in powdered infant formula (PIF). This technical meeting also aimed to provide input to Codex for the revision of the Recommended International Code of Hygienic Practice for Food for Infants and Children. A review of E. sakazakii infections worldwide expanded the findings of the 2004 meeting. While noting that E. sakazakii has caused invasive infection in all age groups, the meeting reiterated the findings of the 2004 FAO/WHO meeting that infants appear to be the group at particular risk. It also highlighted that neonates and also those infants less than 2 months of age are at greatest risk. An analysis of 45 cases indicated that there appears to be two distinct infant groups in terms of the syndrome they develop as a result of an E. sakazakii infection premature infants who develop bacteraemia outside of the neonatal period with most cases occurring in infants less than 2 months of age and term infants who develop meningitis during the neonatal period. This difference in timing of infection may however be related to differences in timing of exposure to E. sakazakii rather than differences in susceptibility but it was also noted that any infant could develop either syndrome at any age. In considering the report of a recent outbreak investigation of nine cases and two deaths associated with the consumption of E. sakazakii contaminated PIF, the meeting noted the value of coordinated disease surveillance and rapid response. The combined efforts of public health and regulatory officials and manufacturers were considered to be an important aspect of the management of the risks associated with E. sakazakii and Salmonella in PIF. Based on the research findings reported since 2004, the meeting noted that E. sakazakii is now recognized as a genotypically and phenotypically diverse bacterial species. Further studies are underway to determine optimal methods of isolation and identification and to address the virulence and the ecology of this pathogen. The meeting reviewed a risk assessment model that had been developed to describe the factors leading to E. sakazakii infection in infants and identify potential risk mitigation strategies. The risk assessment model enables this by facilitating the comparison of different levels of product contamination and of different preparation, handling and feeding scenarios. In addition it provides the means to evaluate microbiological criteria and sampling plans in terms of the risk reductions achieved and the percentage of product lot rejected. The risk assessment model estimates the relative risk of illness from E. sakazakii posed to infants from intrinsically contaminated PIF but does not consider contamination or recontamination from the environment or other sources post-manufacture. After review and discussion of the quantitative risk assessment model on E. sakazakii in PIF, the technical meeting found it to be accurate and valid, based on the approach taken, the assumptions made and the interpretation of data. New data from industry and on consumer and hospital practices related to the use of PIF were incorporated into the model, which was then used to evaluate different risk reduction scenarios. Although the risk assessment does not model specific industry practices, it can evaluate the impact of different levels of contamination in the dry PIF product before use. As requested by Codex, the meeting used the risk assessment model to evaluate the relative risk reduction associated with different risk management interventions for E. sakazakii. Preparation,

13 xiii storage, and handling practices for PIF were addressed as a series of scenarios. One scenario was selected as the baseline, and others were compared to determine their relative risk to the baseline. In general, the use of 70 o C mixing water to reconstitute PIF significantly reduced risk. In other scenarios, reconstitution of PIF with 40 and 50 o C mixing water (unless consumed immediately), holding bottles at room temperature and/or long feeding periods were associated with increased relative risk. The meeting did acknowledge that not all PIF products were formulated to be mixed at 70 o C and recognized that, if recommended, the use of hot water may require specific labelling. The meeting concluded that some of the current preparation instructions on PIF product labels, and those recommended by health authorities may lead to increased risk of E. sakazakii illnesses, and that these should be reviewed in light of the risk assessment results. The meeting also considered that the assessment could be used by FAO/WHO to develop guidance, as requested by the World Health Assembly (WHA) in 2005, on the preparation, use, handling, and storage of infant formula so as to minimize risks where infants cannot be or are not fed breast milk. The risk assessment model was also used to evaluate risk reductions associated with the implementation of microbiological criteria and related sampling plans. The model was found to be a valuable tool to assess and compare a series of criteria and sampling plans and their impact on risk reduction and on the amount of product rejected compared to a situation where no sampling plan was implemented. The meeting concluded that this tool could be applied for E. sakazakii and/or Enterobacteriaceae by risk managers within Codex and FAO and WHO member countries. While the risk assessment model can not be used to evaluate the impact of specific industry practices, such as wet and dry cleaning processes, the meeting did consider the available information on industry practices. The meeting noted that data available from the PIF industry indicates a reduction of both E. sakazakii and Salmonella in PIF through strict separation of the wet and dry phases of product manufacture, and the use of dry cleaning procedures or at least minimising the amount of water used for cleaning procedures whenever possible in specific segments of the manufacturing environment. However, it was noted that it may not be feasible to eliminate the use of water when producing some products, e.g. hypoallergenic powdered infant formulae. Salmonella has been considered a hazard in PIF and related milk-based products for several decades. A review of recent and historical outbreaks of salmonellosis linked to intrinsically contaminated PIF, confirmed its public health importance and the need for risk management vigilance. Outbreaks were only identified by laboratory-based foodborne disease surveillance systems and usually by some unique quality of the epidemic strain, e.g. a rare serotype or the ability to ferment lactose. The meeting recognized that salmonellosis due to contaminated PIF was probably under-reported. The meeting agreed with the 2004 workshop that current microbiological criteria for Salmonella in PIF are considered to be adequate. The meeting further concluded that a risk assessment for Salmonella in PIF was possible. However, it considered that the extensive information currently available on Salmonella and its behaviour provided a good basis for risk management decisions and that there was currently no obvious need for a quantitative risk assessment on Salmonella in PIF.

14 xiv RECOMMENDATIONS In considering the recommendations that would be made the meeting firstly re-endorsed those that had been made by the 2004 FAO/WHO meeting on this issue. The expert meeting made the following additional recommendations. To member countries, FAO and WHO: Develop prevention strategies for E. sakazakii infections caused by contaminated PIF that address the different stages of production and preparation and use of PIF, taking into consideration the risk to infants both within and beyond the neonatal period and of any immune status. Develop educational messages on the safe handling, storage and use of powdered infant formula, including on the health hazards of inappropriate preparation and use, and target health care workers, in both hospitals and in communities, parents, and other caregivers to infants, since E. sakazakii infections have occurred in both hospital and home settings. Review and revise product labels, as appropriate, to provide information to enable caregivers to handle, store, and use the product safely, and on the health hazards of inappropriate preparation. Encourage member countries to establish surveillance and rapid response networks, and facilitate the coordinated investigation by clinicians, laboratorians, public health and regulatory officials to enable the timely recognition and cessation of outbreaks of illness associated with E. sakazakii and the identification of contaminated PIF. Encourage countries to enhance laboratory-based surveillance, including reporting to Salm- Surv, the WHO salmonellosis world wide surveillance network, since laboratory-based surveillance is the only way in which past outbreaks of salmonellosis associated with intrinsically contaminated PIF have been recognized. Encourage laboratories conducting surveillance for Salmonella, and manufacturers and regulators testing for Salmonella in PIF, to use isolation and diagnostic methods which can identify lactose-fermenting strains of this organism, since these have been the cause of some of the outbreaks of salmonellosis associated with PIF. Encourage scientists to determine the optimal isolation and identification methods for E. sakazakii, taking into account the new research data demonstrating genetic and phenotypic diversity in the species. Encourage research to determine ecological niches and virulence factors for E. sakazakii to better target risk mitigation strategies and control measures. Develop and review international guidelines, as requested by the 2005 WHA, educational messages, and product labels regarding the preparation, storage and handling of PIF, considering the results of the E. sakazakii risk assessment model presented in this report. At this time, the meeting did not recommend that FAO/WHO conduct a quantitative risk assessment for Salmonella in PIF. To Codex: Make risk management recommendations based on the outputs of the JEMRA risk assessment for E. sakazakii in PIF and the estimates of potential risk reductions of proposed control measures. In particular, give consideration to the elaboration of sampling plans and

15 xv microbiological criteria for E. sakazakii and Enterobacteriaceae in PIF, and labelling recommendations for PIF, specifically in the revision of CAC/RCP To member countries: Apply the risk assessment in developing national risk management strategies for the reduction of risks associated with PIF, such as appropriate educational programmes. Encourage industry to effectively implement preventive measures and to strengthen those measures that further minimize entry of the micro-organisms of concern to, and avoid their multiplication in the manufacturing environment. To industry Effectively implement preventive measures, including the strengthening of those measures that further minimize entry of the micro-organisms and avoid their multiplication such as the exclusion of water from the processing environment to the extent possible and feasible. The most effective means of achieving the latter is considered to be the implementation of systematic dry-cleaning. Support research that allows further evaluation of the effectiveness of Enterobacteriaceae as an indicator organism that is indicative of conditions in the manufacturing environment or the final product that have an increased potential for harbouring E. sakazakii and/or Salmonella. To FAO and WHO In future calls for data provide more specific details with regard to the type and format of data that is needed in order to enable data providers to target their efforts towards the provision of data which can be effectively used in the risk assessment process.

16

17 Enterobacter sakazakii and Salmonella in powdered infant formula 1 1. Introduction and background In working to ensure the safety and availability of food for infants and young children, one of the more vulnerable groups in our society, the Food and Agriculture Organization of the United Nations and the World Health Organization began addressing the issue of Enterobacter sakazakii and other micro-organisms in powdered infant formula (PIF) in February 2004 through the implementation of a joint technical meeting (FAO/WHO, 2004). Follow-up activities to that meeting included the provision of information and guidance on the risks associated with E. sakazakii and other micro-organisms in PIF to Codex and member countries and the elaboration of a risk assessment tool to enable the evaluation of a range on intervention strategies on risk. In 2005 the World Health Assembly adopted a resolution requesting WHO, in collaboration with FAO to develop international guidelines, educational messages, and product labels regarding the preparation, storage and handling of PIF. In response to these developments a second meeting was implemented on January This focused specifically on E. sakazakii and Salmonella in PIF and addressed, in particular, a number of questions on these pathogens posed by the 37 th session of the Codex Committee on Food Hygiene (CCFH) (14 19 March 2005), the associated risk and means of evaluating control measures. This report documents the discussions and recommendations of that meeting. The meeting was chaired by Mr Aamir Fazil, Public Health Agency of Canada. A total of 16 experts from 11 countries participated in the meeting in their independent capacities and not as representatives of their governments, employers or institutions. These included one expert from the infant formula industry and two academics whose research is being funded by the industry. While these experts participated in the general discussions and exchange of information they did not participate in the final adoption of the conclusions and recommendations of the meeting Background As noted in the first FAO/WHO report on this issue (FAO/WHO, 2004) the 35 th session of the CCFH, following the review of a risk profile, began to address the concerns of member counties about the occurrence of E. sakazakii infections among infants due to consumption of contaminated PIF. The risk profile had concluded that E. sakazakii was emerging as a hazard in PIF causing infrequent but often severe infections in infants, particularly low-birth-weight and/or immunocompromised neonates in the hospital setting. Other Enterobacteriaceae, particularly Salmonella, were also documented as causing disease in infants through low level contamination of PIF. The 35 th Session of CCFH undertook to revise the Recommended International Code of Hygienic Practice for Foods for Infants and Children (CAC/RCP ) and created a Codex Working Group led by the Canadian delegation to begin this work. To support development of the revised code (to be renamed as the Code of hygienic practice for powdered formulae for infants and young children ), the CCFH also asked for expert advice from FAO/WHO. The FAO/WHO meeting convened on 2-5 February, 2004 (FAO/WHO, 2004) confirmed and extended the conclusions of the risk profile considered by the CCFH, and produced an initial risk assessment, which was considered preliminary in nature. A more complex risk assessment model was initiated during that session. That meeting made a number of recommendations which were summarised as follows: To FAO, WHO, Codex, their member countries, NGOs, and the scientific community. In situations where infants are not breastfed, caregivers, particularly of infants at high risk, should be regularly alerted that PIF is not a sterile product and can be

18 2 Introduction and background contaminated with pathogens that can cause serious illness: they should be provided with information that can reduce the risk. In situations where infants are not breastfed, caregivers of high-risk infants should be encouraged to use, whenever possible and appropriate, commercially sterile liquid formula or formula which has undergone an effective point-of-use decontamination procedure (e.g. use of boiling water to reconstitute or by heating reconstituted formula). Guidelines should be developed for the preparation, use, and handling of infant formula to minimize risk. Risk communication, training, labelling and educational activities and approaches to ensure awareness on the issue and appropriate point of use procedures for preparation, storage and use of infant formula should be enhanced or developed. The infant food industry should be encouraged to develop a greater range of commercially sterile formula products for high-risk groups. The infant food industry should be encouraged to reduce the concentration and prevalence of E. sakazakii in both the manufacturing environment and PIF. To this end, the infant food industry should consider implementing an effective environmental monitoring programme and the use of Enterobacteriaceae rather than coliform testing as an indicator of hygienic control in factory production lines. In revising its code of practice, Codex should better address the microbiological risks of PIF, and establish appropriate microbiological specifications for E. sakazakii in PIF. FAO/WHO should address the particular needs of some developing countries and establish effective measures to minimize risk in situations where breast milk substitutes may be used in exceptionally difficult circumstances, e.g. feeding infants of HIV-positive mothers or low-birth-weight infants, lack of refrigerated storage. The use of internationally validated detection and molecular typing methods for E. sakazakii and other relevant micro-organisms should be promoted. Investigation and reporting of sources and vehicles, including PIF, of infection by E. sakazakii and other Enterobacteriaceae should be encouraged. This could include the establishment of a laboratory-based network. Research should be promoted to gain a better understanding of the ecology, taxonomy, virulence and other characteristics of E. sakazakii and of ways to reduce levels in reconstituted PIF. The more complex risk assessment model was further developed by JEMRA consultants and was presented to the Codex Working Group in November That group identified Handling and Storage of Powdered Infant Formula and the Associated Risks of E. sakazakii Infection in Infants as a particular area of concern, given the low-level presence of pathogens in this product. It was acknowledged that the risk assessment model might also be applied to address the impact of the method of manufacture on risk as well as the risk reduction achieved through the implementation of different microbiological criteria, sampling and testing programs. This JEMRA work (CXFH 05/37/9) as well as a draft revised code were presented at the 37 th Session of CCFH, March 2005 in Buenos Aires, Argentina. After review and discussion of the draft revised code, the 37 th Session of CCFH asked FAO and WHO for additional information and scientific advice and requested FAO/WHO to convene an Expert Meeting to look at the following issues:

19 Enterobacter sakazakii and Salmonella in powdered infant formula 3 i. Taking into consideration any existing and new information on E. sakazakii and existing and new data on Salmonella 1, identify if possible the distribution of cases linked to the different types of powdered formula 2 as a function of age, and define specifically the age groups and other groups of infants and young children at greatest risk. ii. Review the dose-response and growth models of E. sakazakii, using new data that is becoming available. iii. Evaluate specific control measures for different manufacturing operations (depending on data provided by manufacturers of powdered formula), which could minimise product contamination by E. sakazakii and evaluate how microbiological criteria for Enterobacteriaceae can be used as an indication of process hygiene. iv. a) In light of new data submitted by ISDI/industry request that the risk assessment be updated to take into consideration this new information and make the output available to the Working Group (in charge of redrafting the proposed draft Code see paragraph 56) for the development of microbiological criteria; b) Use the risk assessment to evaluate the risk reduction associated with various control measures, microbiological criteria and sampling plans. v. Request that the aspects of the risk assessment model addressing preparation, storage and handling of powdered formula be revisited to ensure that all currently used preparation procedures are evaluated. In May 2005, the World Health Assembly (WHA) passed a resolution, WHA58.32, which acknowledged the outcome of the JEMRA and the CCFH meetings. Among other things, the resolution urged member states to take certain protective and preventive actions to minimise the exposure of infants and young children to any hazards associated with PIF such as training and information for caregivers to infants and working with manufacturers to continue to reduce hazards in PIF. The WHA resolution also requested Codex to urgently complete its work related to PIF standard setting, and requested WHO, with FAO to develop guidelines for health care providers on preparation, use, handling, and storage of infant formula and to support independently reviewed research in the area of pathogens in PIF. Outbreaks and sporadic cases of illness and deaths in infants, due to E. sakazakii and Salmonella contaminated PIF are probably under-reported. Nevertheless, reports of cases and outbreaks have continued through 2004 and In late 2004, consumption of E. sakazakii contaminated PIF was linked to nine infections with two deaths in an outbreak among infants in France (Coignard and Vaillant, 2006; Coignard et al., 2006) and five infections with one death among infants in New Zealand (Ministry of Health, New Zealand, 2005). In 2005, Salmonella Agona linked to PIF caused an outbreak among infants in France (InVS, 2005). 1.2 Objectives To respond in a timely way to the ongoing public health concerns regarding E. sakazakii and Salmonella in PIF, the meeting aimed to: Review the risk assessment model for accuracy and validity in terms of the approach taken, the assumptions made, the interpretation of data etc. Review any new epidemiological data that have become available, specifically regarding infants at greatest risk of E. sakazakii infection. 1 The need for any risk assessment work on Salmonella will be reviewed following an initial literature review and consideration of available data. 2 Powdered formula is used here to describe powdered infant formula, follow-up formula, formula for Special Medical Purposes (FSMP) intended for infant, and human milk fortifiers, as described in Section 6.1 of the 2004 Meeting Report Enterobacter sakazakii and other microorganisms in powdered infant formula (FAO/WHO, 2004).

20 4 Introduction and background Review any new data that have become available in relation to Salmonella in PIF and if necessary revise the recommendations of the previous meeting with regard to this pathogen. Using the risk assessment model, estimate the potential risk reduction associated with the implementation of various microbiological criteria and their associated sampling plans for PIF production. Specifically address E. sakazakii in the product and Enterobacteriaceae as an indicator of process control. Using the risk assessment model, estimate the potential risk reduction of control measures implemented during reconstitution, storage, and handling/feeding of PIF. Based on the risk assessment model, and other relevant data, provide a response to the issues raised by the 37 th session of the CCFH, as outlined above. 1.3 Scope This meeting builds on the work done by the 2004 expert meeting (FAO/WHO, 2004) which recommended, among other things, that JEMRA expand and complete the more extensive risk assessment initiated at that meeting. Thus, the scope of this meeting was to address the risk associated with PIF intrinsically contaminated with E. sakazakii and Salmonella and in doing so respond to the specific questions posed by CCFH. The meeting reviewed the risk assessment and applied it as a tool to provide estimates of relative risk reductions associated with different levels of contamination of PIF, different preparation and use scenarios, and a range of microbiological criteria and their associated sampling plans. In this manner the meeting addressed the requests of CCFH for scientific advice to be used in revising the current Recommended International Code of Hygienic Practice for Food for Infants and Children. Within the context of this work PIF was considered to include products in powdered form specially manufactured and presented to be used by infants, either as a breast-milk substitute after preparation with water, or to modify prepared breast-milk substitutes or to fortify human milk. Included products are, therefore, infant formula (as defined in Codex Stan ), follow-up formula (as defined in Codex Stan ), formula for special medical purposes intended for infants (as defined in Codex Alinorm 04/27/26, Appendix V 5 ), formulas for special medical purposes for the partial feeding of infants (covered by Codex Stan ) and human milk fortifiers as defined in the 2004 meeting (FAO/WHO, 2004). Throughout the report the term powdered infant formula or PIF includes all the products mentioned above, but does not include cereals or other any other product added to breast milk or used as breast milk substitutes that are not specifically manufactured for that purpose. This expert meeting reviewed the most recent available epidemiological and microbiological science related to E. sakazakii and Salmonella and their intrinsic contamination of PIF. The scientific review specifically addressed susceptible human populations, information from recent outbreaks of illness, new diagnostic assays, virulence and ecological factors of the organisms. The meeting also updated and/or expanded on previous information on industry and consumer practices relevant to E. sakazakii and Salmonella contaminated PIF. 1.4 Micro-organisms other than Enterobacter sakazakii and Salmonella The 2004 meeting addressed the potential risk of a range of micro-organisms in PIF and reported clear evidence of causality (Category A) for E. sakazakii and Salmonella. Other micro-organisms were considered and assigned to one of the two other categories based on 3 Available at: 4 Available at: 5 Available at: 6 Available at:

21 Enterobacter sakazakii and Salmonella in powdered infant formula 5 the strength of the evidence of a causal association between their presence in PIF and illness in infants. This meeting agreed to retain the categorization scheme developed at the 2004 meeting. No outbreaks associated with the organisms in category B linked to their occurrence in PIF have been reported that would warrant considering the re-assignment of those organisms to category A. Hence the only organisms in Category A Clear evidence of causality remain Salmonella enterica and E. sakazakii. It was noted that some organisms which have been isolated from PIF and are associated with neonatal infections had not been included in Category B Causality plausible, but not yet demonstrated. These include Escherichia coli, Serratia spp. and Acinetobacter spp. (Kaufman and Fairchild, 2004; Lee et al., 2004; Estuningsih et al., 2006). This is information that had not been published before the 2004 meeting and therefore these organisms should now be included in Category B. Similarly with respect to Category C Causality less plausible or not yet demonstrated the omission of coagulase-negative staphylococci which cause more neonatal infections in very low birth weight neonates than Staphylococcus aureus (already in Category C) was noted (cf. Kaufman and Fairchild, 2004; Lee et al., 2004). In addition, in 2005 there were two papers reporting the case of a five month old infant with botulism in 2001 and a possible link to PIF (Brett et al., 2005; Johnson et al., 2005). Clostridium botulinum was isolated from the baby s faeces, foodstuffs in the house and opened and unopened PIF of the same batch. Whilst C. botulinum type A was isolated from both the baby s faeces and the opened formula, the unopened formula contained C. botulinum type B that did not match the type isolated from the baby. It was considered that this single case did not warrant re-categorisation of C. botulinum. No outbreaks due to the remaining organisms in Category C have been reported and hence these organisms remain in this category. The updated categorization is presented in Table 1. Table 1. Categorization of the micro-organisms or microbial toxins of concern in powdered infant formula based on the strength of evidence of a causal association between their presence in PIF and illness in infants. 7 Category Category A organisms clear evidence of causality Category B organisms causality plausible, but not yet demonstrated Category C organisms causality less plausible or not yet demonstrated Organisms included Enterobacter sakazakii, Salmonella enterica Pantoea agglomerans and Escherichia vulneris (both formally known as Enterobacter agglomerans), Hafnia alvei, Klebsiella pneumoniae, Citrobacter koseri, Citrobacter freundii, Klebsiella oxytoca, Enterobacter cloacae, Escherichia coli, Serratia spp. and Acinetobacter spp. Bacillus cereus, Clostridium difficile, Clostridium perfringens, Clostridium botulinum, Listeria monocytogenes, Staphylococcus aureus and coagulase-negative staphylococci. 7 This categorization was developed by the FAO/WHO expert meeting on Enterobacter sakazakii and other microorganisms in powdered infant formula, 2 5 February 2004 (FAO/WHO, 2004).

22

23 Enterobacter sakazakii and Salmonella in powdered infant formula 7 2. Epidemiological and microbiological aspects of Enterobacter sakazakii 2.1 Epidemiology and Public Health Although E. sakazakii has caused illness in all age groups, certain groups are likely to experience higher disease rates. As reported in the first meeting (FAO/WHO, 2004) a United States FoodNet survey in 2002 estimated the annual rate of invasive E. sakazakii infection (based on isolation of the organism from sterile sites only) to be 1 per infants (i.e., children <12 months of age) whereas the rate among low-birth-weight neonates was 8.7 per in the nine FoodNet surveillance sites (C. Braden, personal communication, 2004). Similarly, the annual incidence of invasive E. sakazakii infection was estimated at 9.4 per 100,000 infants of very low-birth-weight (<1500 g) in a study conducted in 19 neonatal intensive care units (Stoll et al., 2004). Although the incidence of illness is not known for older paediatric or adult populations, worldwide, published reports of bloodstream or central nervous system infections are rare: there are three reported cases among children aged >12 months (Arseni et al., 1987; Lai, 2001), and eight in adults, six among persons aged >70 years (Lai, 2001). The estimated incidence of E. sakazakii bloodstream infection among mixed ages in England, Wales and Northern Ireland ranged from 0.09 to 0.12/100,000 population between , based on 2001 census data (HPA, 2004). Therefore, children >12 months and adults are assumed to be at lower risk than infants for invasive infection with E. sakazakii. Among infants, those born prematurely or with low-birth-weight (<2500 g) are reported to be at highest risk for severe infection (Lai, 2001; FAO/WHO, 2004; Gurtler, Kornacki and Beuchat, 2005). However, in a review of 45 cases of bloodstream and central nervous system infections abstracted from the English language literature (n=32) and CDC case consultations (n=13), this group of infants appears, on average, to develop isolated E. sakazakii bacteraemia more often than meningitis (Table 2). A list of the cases on which this analysis is based is provided in Appendix E. While PIF was established as the source of E. sakazakii in some of the cases, in many cases PIF was neither epidemiologically nor microbiologically implicated as the source of infection. However, in such cases no other source of infection has been epidemiologically or microbiologically implicated. The analysis indicated that premature infants tend to develop infection at a later median age, 35 days of life, than the infants who develop meningitis. According to recent CDC case reports bacteraemia cases have occurred as late as age 10 months in an immunosuppressed infant, and 8 months in a previously healthy infant (A. Bowen, personal communication, 2006). In contrast, meningitis appears to occur more frequently among infants born following term gestation and at a birth weight approaching the normal range. However, these infants also generally develop symptoms before the age of one week. Similar proportions of infants with bacteraemia (27%) and meningitis (33%) develop symptoms outside of the hospital setting. Infants are particularly susceptible to meningitis during the first several weeks of life, and may be more likely to develop meningitis rather than isolated bacteraemia if exposed to E. sakazakii during this window. Consequently, although there appear to be two distinct infant risk groups, namely, premature infants developing isolated bacteraemia after one month of age, and term infants developing meningitis during the neonatal period, the difference in timing of infection may be related to differences in timing of exposure to E. sakazakii rather than differences in susceptibility. Outcomes differ between types of infection (Table 3). Infants with bacteraemia tend to fare better; mortality among these cases is 10%. Conversely, those with meningitis frequently develop complications including seizures and brain abscesses, infarcts, or cystic changes. Of

24 8 Epidemiological and microbiological aspects of Enterobacter sakazakii the infants with meningitis in this series 44% died and the majority of survivors experienced long-term neurological consequences. Table 2. Characteristics of infants who develop different syndromes due to E. sakazakii infection Characteristics of infants Bacteraemia N=12 Meningitis N=33 Median Range Median Range Birth weight, g 850 [ ] 2454 [ ] Estimated gestational age, weeks 27.8 [ ] 37 [30 40] Age at onset, days 35 [7 300] 6 [2 35] Table 3. Outcomes of E. sakazakii infection in infants according to syndrome Outcome Bacteraemia Meningitis n/n % n/n % Seizures 1/ /33 33 Brain abscess or cyst 0 0 7/33 21 Hydrocephalus* 0 0 8/19 42 Developmental delay* /19 53 Mortality 1/ /32 44 *Outcome among survivors In summary, although E. sakazakii has caused invasive infection in all age groups, the meeting reiterated the findings of the 2004 FAO/WHO meeting that infants appear to be the group at particular risk with neonates and also those less than 2 months of age at greatest risk. E. sakazakii meningitis tends to develop in infants during the neonatal period, which mirrors age-based meningitis rates due to other pathogens, and which may be due to immunologic or neurovascular processes in this age group (Schuchat et al., 1997). E. sakazakii bacteraemia tends to develop in premature infants outside of the neonatal period with most cases occurring in infants less than 2 months of age. However, infants with immunocompromising conditions have developed bloodstream infections as late as age 10 months and previously healthy infants have also developed invasive disease outside the neonatal period. Infections have occurred in both hospital and outpatient settings and it was noted that as older infants generally live at home in the community, infections in such infants may be more likely to be under-reported Outbreaks Update since 2003 In 2004, a small outbreak occurred in New Zealand which was linked to PIF used in a nursery. Subsequently, a premature infant died after contracting E. sakazakii meningitis. A follow-up investigation in the neonatal intensive care unit (NICU) found that four other babies had been colonised with the organism. The PIF, which was implicated, was confirmed microbiologically. As a result of this outbreak, there was an inquiry into actions of sector agencies in relation to contamination of infant formula with E. sakazakii, and E. sakazakii invasive disease was made notifiable in New Zealand in 2005 (Ministry of Health, New Zealand, 2005). Another outbreak due to E. sakazakii occurred in France in 2004 (Coignard and Vaillant, 2006; Coignard et al., 2006). A total of nine cases were reported, with two deaths. Syndromes included fatal meningitis (2), conjunctivitis (1), hemorrhagic colitis (1) and colonization (5). All infants were premature and under 2000g (low birth-weight), except for